Hugging Face's logo

Datasets:
inference-optimization
/
RazorCal

Tasks:
Text Generation
Modalities:
Text
Formats:
parquet
Languages:
English
Chinese
Size:
1K - 10K
ArXiv:
Tags:
calibration
model-compression
pruning
mixture-of-experts
Mixture of Experts
arxiv:2609.30465
Libraries:
Datasets
pandas
Polars
License:
Dataset card Data Studio Files
xet
 Community
RazorCal / README.md
kylesayrs's picture
kylesayrs
Add dataset card
696498e verified
preview code
|
Raw History Blame Contribute Delete
9.06 kB
metadata 
license: other
license_name: source-specific-terms
license_link: >-
  https://huggingface.co/datasets/inference-optimization/RazorCal/blob/main/LICENSE-DATA
language:
  - en
  - zh
size_categories:
  - 1K<n<10K
task_categories:
  - text-generation
tags:
  - calibration
  - model-compression
  - pruning
  - mixture-of-experts
  - moe
  - arxiv:2609.30465
dataset_info:
  features:
    - name: domain
      dtype: string
    - name: sub_source
      dtype: string
    - name: coding_subclass
      dtype: string
    - name: messages
      list:
        - name: role
          dtype: string
        - name: content
          dtype: string
        - name: tool_calls
          list:
            - name: id
              dtype: string
            - name: type
              dtype: string
            - name: index
              dtype: int64
            - name: function
              struct:
                - name: name
                  dtype: string
                - name: arguments
                  dtype: string
    - name: tools
      list:
        - name: type
          dtype: string
        - name: function
          struct:
            - name: name
              dtype: string
            - name: description
              dtype: string
            - name: parameters
              dtype: string
  splits:
    - name: train
      num_bytes: 7506600
      num_examples: 2048
  download_size: 7414027
  dataset_size: 7506600
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*

RazorCal

RazorCal is a general-purpose calibration corpus released with RAZOR. It contains 2,048 samples across seven domains (about 8 MB). Each record includes input messages and source attribution. The records carry no architecture-specific fields, so the corpus also suits FFN or layer pruning, perplexity measurement and other calibration-based compression; RAZOR uses it to estimate per-expert statistics. Reasoning traces are omitted because the calibration loader does not consume them. Potential contact identifiers, home-directory usernames, private-network examples and literal password examples are replaced with placeholders.

This repository re-publishes the corpus in a row-oriented format: each of the 2,048 samples is one dataset row. It is converted from the original single-file release at Nickyang/RazorCal, which is also included unchanged at source/RazorCal.json so that the conversion can be verified against the original artifact.

Composition

Domain Samples
Coding 512
Math 256
Science (STEM) 256
Chinese-STEM 256
Instruction Following 256
Tool Calling 256
World Knowledge 256
Total 2,048

Format

Each row is one sample with the following columns:

Column Type Description
domain string one of the seven domains above
sub_source string recorded upstream source attribution
coding_subclass string, nullable subclass label, present on Coding samples
messages list of structs chat messages with role, content and optional tool_calls (id, type, index, function.name, function.arguments)
tools list of structs, nullable tool definitions with type and function (name, description, parameters)

The corpus defines a single train split holding all 2,048 samples; no train/validation/test semantics are implied.

Conversions applied to the original file, needed to keep a single valid arrow schema:

  • tool_calls[].function.arguments and tools[].function.parameters are free-form JSON objects whose value types conflict across records (for example a string and a number under the same key path of different records). They are recorded as JSON strings; parse them with json.loads before use.
  • The optional message fields tool_call_id and name are null for every record in the original file and are omitted here.
  • Empty tool_calls and tools lists are recorded as null.

Everything else is preserved verbatim. The corpus metadata recorded in the original file's meta object:

{
  "name": "RazorCal",
  "total_samples": 2048,
  "max_length_per_sample": 32000,
  "domains": {
    "Math": 256,
    "Science (STEM)": 256,
    "Chinese-STEM": 256,
    "Instruction Following": 256,
    "Tool Calling": 256,
    "World Knowledge": 256,
    "Coding": 512
  },
  "coding_subclass_distribution": {
    "swe_issue_fix": 40,
    "engineering_oop_api": 111,
    "cp_python_stdin": 138,
    "cp_cpp_stdin": 99,
    "func_signature_completion": 100,
    "leetcode_class_solution": 24
  },
  "homepage": "https://github.com/nick7nlp/RAZOR",
  "description": "RazorCal is a multi-domain calibration corpus released with RAZOR. It is a general-purpose calibration mixture: the records carry no architecture-specific fields and can be used for expert, FFN or layer pruning as well as other calibration-based compression. Source labels record attribution; see the datasheet for provenance, privacy processing, and usage limitations.",
  "license": "Source-specific terms apply; see LICENSE-DATA.",
  "privacy_processing": "Reasoning traces are omitted. Contact identifiers, home-directory usernames, private-network examples, and literal password examples are replaced with placeholders."
}

Loading the rows:

import json

from datasets import load_dataset

dataset = load_dataset("inference-optimization/RazorCal", split="train")

sample = dataset[0]
messages = sample["messages"]
for call in messages[-1]["tool_calls"] or []:
    call["function"]["arguments"] = json.loads(call["function"]["arguments"])

For the RAZOR CLI, use the unchanged original file:

curl -L -o RazorCal.json \
  https://huggingface.co/datasets/inference-optimization/RazorCal/resolve/main/source/RazorCal.json

razor saliency --model <model-path> --data RazorCal.json --out out/sal

Recorded sources

A source label records attribution information; it is not, by itself, proof of a record-level match to an upstream release.

Domain sub_source labels and counts Recorded source description
Math nemotron-cascade2-math (160); nemotron-math-v2-high-aime-style (96) Nemotron-Cascade-2-SFT-Data math slice; Nemotron-Math-V2
Science (STEM) Nemotron-Science-V1-RQA (176); Nemotron-Science-V1-MCQ (80) Nemotron-Science-V1
World Knowledge OpenHermes-2.5 (256) OpenHermes-2.5 via smoltalk2
Coding Dolci-Python-Algorithms (205); competitive-programming-python (141); competitive-programming-cpp (115); Nemotron-SFT-SWE-V2-agentless (40); Nemotron-OpenCode-general (11) Dolci; competitive-programming, SWE and OpenCode corpora
Instruction Following Dolci-Precise-IF (256) Dolci instruction-following corpus
Tool Calling Nemotron-Agentic-tool-calling-v2 (256) Nemotron agentic/tool-calling corpus
Chinese-STEM multilingual-chinese-math (86); multilingual-chinese-code (85); multilingual-chinese-stem (85) Multilingual Chinese categories

The nemotron-cascade2-math records were taken as a whole domain slice rather than sampled individually, so the label records the upstream dataset but not a per-record identifier within it. Source labels alone do not establish provenance. See FIDELITY_AUDIT.md for provenance boundaries and LICENSE-DATA for recorded license information.

Usage limitations

RazorCal is calibration data, not an independent evaluation benchmark. Domain coverage does not guarantee preservation of model quality. Review possible overlap with evaluation data for the intended use.

The original record array contains 587 tool messages without populated tool_call_id links and 11 tool definitions with empty parameters objects; in this repository the always-null tool_call_id and name fields are omitted, as documented under Format. Tokenizer chat templates and schema validators may require additional handling at load time.

Public source material can still contain personal references or unsuitable content; source labels do not establish that records are free of such content. See DATASHEET.md for intended use and limitations.

Citation and licensing

Source-specific terms continue to apply; no single replacement license is asserted for all RazorCal records. Retain applicable upstream attribution. See LICENSE-DATA and NOTICE before use or redistribution.

This repository contains no code; LICENSE-CODE is retained from the upstream release for completeness. See DATASHEET.md for intended use and limitations and FIDELITY_AUDIT.md for the scope of available provenance evidence.

@misc{song2026razorpruningreplaceableexperts,
      title={RAZOR: Pruning Replaceable Experts in LLMs},
      author={Mingyang Song and Mao Zheng},
      year={2026},
      eprint={2609.30465},
      archivePrefix={arXiv},
      primaryClass={cs.LG},
      url={https://arxiv.org/abs/2609.30465},
}